The RAS/RAF/MEK/ERK MAPK pathway is a key signaling pathway involved in the regulation of normal cell proliferation, survival, and differentiation [1
]. Under normal circumstances, the serine/threonine kinase BRAF is activated by NRAS (neuroblastoma RAS viral oncogene homolog) [2
] and in turn phosphorylates the downstream proteins MEK1/2 (MAP2K, mitogen-activated protein kinase kinase), which then activate ERK1/2 [4
]. Activated ERKs translocate to the nucleus, where they phosphorylate and regulate different transcription factors, which leads to changes in gene expression [5
]. However, common oncogenic mutational activation of NRAS or BRAF is observed in human tumors [6
]. Approximately 50–60% of metastatic melanomas contain an activating mutation in the BRAF oncogene. Of the mutations in BRAF, over 90% affect amino acid position 600, with the vast majority resulting in substitution of a valine into a glutamic acid (BRAFV600E
), but also other substitutions at this position are found [6
]. These genetic alterations result in constitutive activation of the MAPK signaling pathway, which supports cell proliferation and tumor cell growth through several mechanisms, including reduced apoptosis, increased metastatic potential, invasiveness, and immune suppression [6
]. As a consequence of this knowledge, new therapeutic approaches using specific inhibitors as targeted therapy against the MAPK signaling pathway members were developed for the treatment of melanoma patients.
BRAF kinase inhibitors (BRAFi) like vemurafenib (Vem; marketed as Zelboraf) and dabrafenib (Dabra) have become the standard targeted therapy for melanoma patients with BRAF mutations [11
]. Unfortunately, after the first evidence of objective response, most patients developed resistance to BRAFi monotherapies which was manifested by progressive disease and rapid relapse often caused by a reactivation of the MAPK pathway, e.g., appearance of additional NRAS or other MEK-activating mutations [13
To address this problem, specific MEK inhibitors (MEKi) were developed to additionally inhibit the cascade further downstream. Combined treatment with Vem and the MEKi cobimetinib (Cobi) or Dabra and the MEKi trametinib (Tram) resulted in an increase of progression-free survival, compared to BRAFi alone [15
]. Both these combinations of BRAFi/MEKi were recently approved by both the FDA and the European Commission for the treatment of advanced melanoma patients. Nevertheless, secondary resistance develops frequently [15
]. Therefore, several phases II and III trials are currently evaluating triple therapies of BRAF/MEK inhibitor therapy plus checkpoint inhibitor treatment with anti-PD1 (NCT02910700, NCT02967692, NCT02858921, NCT02130466). Other immunological treatment modalities are also being investigated as combination partners.
As a new therapeutic approach to potentially increase survival and delay relapse, these clinically applied BRAFi/MEKi could be combined with cellular immunotherapy, e.g., chimeric antigen receptor (CAR)-T-cell therapy.
Because the MAPK pathway is also involved in immune cell function and survival, BRAFi and MEKi are likely to influence immune functions. Previous studies have shown that BRAF and MEK inhibitors may have an influence on immune cells and can modulate their functions [17
]. The effects of the BRAFi must be explained with a lack of specificity for the mutated BRAF because the non-malignant cells only harbor the wild-type version of this kinase. The MEKi, in contrast, target non-mutated MEK1/2, and could thus possibly interfere with a pathway essential for the activity of immune cells. MEK-inhibitors as monotherapy to treat melanoma patients with MAPK pathway activating mutations other than BRAFV600
are currently being explored [20
Several CARs against different antigens also expressed on melanoma were already tested in clinical trials (NCT03060356, NCT01218867, NCT02107963, NCT02830724). In previous work [21
], we have generated a CAR specific for chondroitin sulfate proteoglycan 4 (CSPG4), also known as melanoma-associated chondroitin sulfate proteoglycan (MCSP), or high molecular weight melanoma-associated antigen (HMW-MAA), which is a cell-surface antigen expressed on 90% of melanoma primary tumors and metastases, but also on sarcomas, astrocytomas, gliomas, and neuroblastomas [22
], and therefore we consider this an ideal target antigen. We have shown that T cells transfected with this CAR mediated effective antigen-specific tumor cell lysis in vitro and in vivo and also induced the secretion of pro-inflammatory cytokines [21
A combination of BRAFi/MEKi treatment with CSPG4-specific CAR-T-cell therapy would be a new and probably more efficient approach for melanoma therapy. To analyze the possible immunological effects of BRAFi/MEKi, we tested in vitro how the application of these kinase inhibitors influences the functionality of CAR-transfected T cells. We studied in detail CAR-T-cell activation, cytokine secretion, and cytolytic capacity, and found differential effects of the two different BRAFi/MEKi combinations on these CAR-T-cell functions.
The findings of this study are highly relevant for the future use of BRAF and MEK inhibitors in combination with adoptive CAR-T-cell therapy or other immunotherapies. Of the two approved BRAFi/MEKi combinations, the Dabra + Tram combination had a much smaller negative effect on CAR-T-cell functionality than the Vem + Cobi combination. Our data provide a clear rationale for the combination of targeted therapy and immunotherapy for melanoma and may further expand the understanding of BRAF and MEK inhibitor effects on the immune system.
The strategy of combining BRAFi and MEKi with immunotherapy requires a better understanding of the effects of kinase inhibition on normal immune cell function. Although the two currently approved combinations of BRAFi and MEKi appear similarly effective against melanoma [28
], their effects on healthy cells, not bearing BRAF mutations, but employing the respective signaling pathway, may significantly differ. BRAFi have for example a paradoxical effect on wild-type BRAF [29
], which is more pronounced for Vem than for Dabra [30
]. The specificity of the MEKi for MEK1 and MEK2 also varies [31
]. Therefore, the different BRAFi and MEKi may differentially interfere with the various effector functions of CAR-transfected T cells. Tumor rejection depends on T-cell activation and subsequent cytokine secretion and lytic activity of these T cells. Thus, in this study, we thoroughly investigated the effects of these inhibitors on activation, cytokine secretion, and the cytolytic capacity of CSPG4-specific CAR-transfected CD8+
T-cell in in vitro assays.
In both the antigen-specific activation of CAR-T cells and the antigen-specific cytokine secretion by CAR-T cells, we observed the mainly inhibitory effects of the BRAFi and MEKi used, which would argue against a combination with CAR-T-cell-based immunotherapy. However, the different inhibitors and their combinations clearly varied in the intensity of these effects: Vem alone, Cobi alone, Tram alone, and Vem + Cobi had the largest negative influence, while Dabra alone had the mildest negative influence. Of note is that of the two clinically used combinations of BRAFi/MEKi, Dabra + Tram had a much smaller negative effect than Vem + Cobi. This was also the case when looking at the lytic capacity of CAR-T cells.
Considering antigen-specific cytokine secretion by CAR-T cells, several observations are important for an intended combined clinical application of BRAFi/MEKi with CAR-T cells, since efficacy as well as toxicity can be influenced:
(i) The pro-inflammatory cytokines IL-2, TNF, and IFNγ are important for a good T-cell response against the tumor. IL-2 promotes the differentiation of T cells into effector and memory T cells [32
], TNF was originally described as anti-tumorigenic [33
], and IFNγ has a number of important functions including macrophage activation, major histocompatibility complex induction, and Th1 differentiation [34
]. However, the downside of an efficient secretion of these cytokines can be a type of systemic inflammatory immune response, which is similar to severe infections and characterized by symptoms like hypotension, pyrexia, tachycardia, headache, swelling, redness, or nausea [35
]. This so-called cytokine release syndrome (CRS) is a feared side effect of CAR-T-cell therapy caused by a massive systemic release of pro-inflammatory cytokines by the transferred cells [36
]. In the serum of patients where CRS was observed, pro-inflammatory cytokines like IL-6, TNFα, and IFNγ were consistently elevated [40
]. CAR-transfected T cells were least compromised in the production of these cytokines in the presence of Dabra alone compared to the other kinase inhibitors. Secretion of these cytokines was reduced by the presence of Dabra + Tram compared to Dabra alone. This might have a positive effect on the reduction of CRS side effects. Since IL-2, TNF, and IFNγ are nevertheless necessary for an anti-tumor response, the use of Dabra + Tram might form a good balance between preventing an exaggerated cytokine release causing CRS on the one hand and a minimum secretion of the cytokines seen in the setting with Vem + Cobi on the other hand.
(ii) Dabrafenib facilitated the CAR-induced secretion of a very high quantity of IL-6, whereas T cells stimulated in the absence of kinase inhibitors did not produce this cytokine. In the presence of the MEKi some IL-6 was also produced antigen-specifically. In the presence of Vem, no IL-6 was detected. Interestingly, some studies have shown that constitutive activation of the MAPK pathway by the BRAFV600E
mutation induces the downstream production of IL-6 [10
]. The release of IL-6 we observed was probably not caused by a paradoxical effect of Dabra, because it would then also be expected with Vem. Therefore, the molecular reasons for this observation remain to be elucidated. IL-6 is a multifunctional cytokine that plays a central role in host defense due to its wide range of immune and hematopoietic activities and its potent ability to induce the acute phase response [42
]. On the other hand, IL-6 plays a central role in CRS [40
]. Due to the fact that CRS is a severe and potentially deadly side effect, the application of Dabra should only be combined with CAR-T-cell therapy together with Tram, because this mitigates the IL-6 secretion and thus also the possible side effects of CRS.
(iii) CAR-transfected T cells did not secrete any IL-4 and secreted only very low quantities of IL-10, which were further reduced by the different BRAFi and MEKi. Since IL-10 is an anti-inflammatory cytokine that supports melanoma cell proliferation and inhibits anti-tumor responses, and as production of IL-4 can induce IL-10 secretion [43
], the production of these anti-inflammatory cytokines after application of CAR-T cells in the patient could lead to an inhibition of tumor-reactive T cells and prevent effective recognition and lysis of cancer cells and even promote melanoma growth and should therefore be avoided.
Others have also tested the effect of therapeutically relevant inhibitor concentrations on the cytolytic capacity of CAR-T cells [26
]. In contrast to our results, Gargett et al. [26
] showed that Vem alone clearly inhibited the cytolytic capacity of these CAR-T cells, and Dabra combined with Tram also inhibited the cytolytic capacity, but to a lesser extent. In line with our results, Dabra alone did not inhibit the lytic activity. It is important to note in this case that the concentrations of Dabra and Tram were chosen at the higher end of the patient plasma range [26
]. Furthermore, Gargett et al. [26
] incubated the CAR-T cells for 48 h in the presence of BRAFi/MEKi without stimulation, and then these cells were co-cultured for 6 h in the presence of these kinase inhibitors with chromium-labeled target cell lines. Moreover, they tested T cells from melanoma patients, whereas we tested T cells from healthy donors. Finally, the observed differences might be explained by the use of a third generation CAR containing CD3ζ, CD28, and OX-40 signaling domains by Gargett and co-workers [26
Our findings not only have consequences for the use of BRAFi/MEKi in combination with CAR-T cells, but also in a more general sense considering immune responses or combinations with other immunotherapies. For example, our finding that the MEK inhibitor Cobi used as a single agent can inhibit the lytic capacity of T cells might be of importance in studies using MEKi without BRAFi in the setting of melanoma with non-mutated BRAF but mutated NRAS [20
]. Anti-tumor T-cell responses could be influenced in such settings. Moreover, other combinations of MAPK-pathway-targeted therapy and immunotherapy were tested for melanoma as well. For example, it was shown that the combination of checkpoint inhibitors with BRAFi and MEKi is reasonable but dangerous and can cause severe side effects [44
]. A combination of Dabra, Tram, and ipilimumab was tested, resulting in colitis followed by intestinal perforation in two out of seven patients [44
], which was caused by Tram, since in the combination of Dabra and ipilimumab, only one grade 3 colitis was observed in 25 patients. It is not clear yet how Tram, or any MEK inhibitor, contributes to the toxicity of ipilimumab [44
]. Other authors tested in a phase I study a combination of a PD-L1-antibody with Dabra and Tram in metastatic melanoma patients with mutated BRAFV600
], and showed that a combination is possible.
Although we have investigated the influence of the BRAFi and MEKi on immunotherapeutical effector functions of the CAR-T cells, the differential effects on the intracellular signaling induced by the CAR remain to be elucidated. The next steps to understand the observed differences should be a thorough analysis of the phosphorylation state of the respective signaling cascade upon CAR stimulation in the presence of the inhibitors. In this context, also the influence on the mechanism of killing should be addressed to distinguish between granzyme/perforin- and death-receptor-mediated killing. Such knowledge will be valuable in understanding the different effects of the different inhibitors, and help in the improvement of such therapies and the design of new small molecule inhibitors of the RAS/RAF/MEK/ERK-pathway.
4. Materials and Methods
4.1. Cell Culture Media
R10 medium is RPMI 1640 (Lonza, Basel, Switzerland) supplemented with final concentrations of 10% (v/v) heat-inactivated fetal bovine serum (PAA, GE Healthcare, Piscataway, NY, USA), 2 mM l-glutamine (Lonza), 100 U/mL penicillin, 100 μg/mL streptomycin (Lonza), 2 mM HEPES (PAA, GE Healthcare), and 20 μM β-mercaptoethanol (Gibco, Life Technologies, Carlsbad, CA, USA).
4.2. BRAF and MEK Inhibitors
All inhibitors were purchased as pure substances: vemurafenib (PLX4032) from Adooq Bioscience, Irvine, CA, USA, cobimetinib (GDC-0973), trametinib (GSK1120212), and dabrafenib (GSK2118436A) from AbMole BioScience, Houston, TX, USA. BRAF and MEK inhibitor concentrations used in our in vitro experiments were based on the description on the package insert of the providers and published serum concentrations [16
]. Used final concentrations are summarized in Table 1
4.3. Cell Lines
CRL-1992™) is a CSPG-negative TAP-deficient TxB hybrid cell line. The CSPG4+
A375M melanoma cell line (ATCC®
CRL-3223™) was described previously by Kozlowski et al. [46
]. Both cell lines were cultured in R10 medium.
4.4. T-Cell Isolation
All human material from healthy volunteers was obtained after written informed consent for inclusion before they participated in the study. The study was conducted in accordance with the Declaration of Helsinki, and the protocol was approved by the institutional review board of the Friedrich-Alexander-Universität Erlangen-Nürnberg (date: 14 September 2016; reference number: 251_16 B). Peripheral blood mononuclear cells (PBMCs) were purified by density centrifugation using the Lymphoprep reagent (Axis-Shield poC AS, Oslo, Norway). CD8+ T cells were isolated by Magnetic Activated Cell Sorting (MACS) using CD8-specific microbeads (Miltenyi Biotech, Bergisch Gladbach, Germany). The CD8+ fraction was cultured in R10 medium supplemented with 10 ng/mL IL-7 at a concentration of 1 × 106 cells/mL. The isolated cells were rested overnight at 37 °C until they were used for further experimental procedures.
4.5. RNA Transfection
The composition of the CSPG4-specific CAR was described previously [21
] and featured a CD28-CD3ζ CAR backbone [47
] 5′ of the IgG-spacer region. The DNA encoding the CARs was inserted into the pGEM4Z-5′UTR-sig-husurvivin-DC.LAMP-3′UTR RNA-production vector [48
] (kindly provided by Kris Thielemans), replacing the sig-husurvivin-DC.LAMP sequence. RNA was produced using the mMESSAGE mMACHINE T7 Ultra kit (Life Technologies, Carlsbad, CA, USA). RNA was purified with an RNeasy Kit (Qiagen, Hilden, Germany). Electroporation of T cells was performed as described in detail previously [49
4.6. Staining of CSPG4-Specific CAR on Transfected T Cells and CSPG4 on Target Cells
For the analysis of CSPG4-specific CAR on transfected T cells, these cells were stained 4 h after electroporation. Detection of CAR-expression was performed by using a goat F(ab’)2 anti-human IgG-RPE antibody (Southern Biotech, Birmingham, AL, USA, CSGP4-expression on the surface of target cell lines was determined using purified mouse-anti-CSPG4 antibody; clone 9.2.27 (BD). The secondary antibody used was PE-conjugated goat-anti-mouse-Ig polyclonal antibody (BD). Expression was measured directly via a FACScan cytometer (BD). Results were evaluated with CellQuest software (BD) and FCS Express software (FCS Express 5 Flow Research Edition) (DeNovo Software, Glendale, CA, USA).
4.7. Staining of T-Cell Activation Markers
T cells were used 4 h after electroporation, and were co-cultured with the target cells A375M or T2 overnight in R10 medium at a 1:1 ratio with 106 cells per mL in total, in the absence or presence of BRAFi/MEKi. The activation markers CD25 and CD69 on T cells were analyzed by flow cytometry. T cells were stained with anti-CD25-FITC and CD69-PE antibodies and surface marker expression was measured directly via FACScan cytometer (BD). T cells were distinguished from target cells by forward/sideward scatter gating and results were evaluated with CellQuest software (BD) and FCS Express software (DeNovo Software). The specific mean fluorescence intensity (MFI) was calculated by subtraction of the background MFI obtained with isotype antibodies by using mouse IgG1 κ isotype control FITC antibody (BD) and mouse IgG1 isotype control PE antibody (Miltenyi Biotech), respectively.
4.8. Cytokine Secretion Assay
T cells were used 4 h after electroporation and were co-cultured with the target cells A375M or T2 overnight in R10 medium at a 1:1 ratio with 106 cells per mL in total, in the absence or presence of BRAFi/MEKi. The cytokine concentrations in the supernatants were determined with the Cytometric Bead Array human Th1/Th2 Cytokine Kit II (BD Bioscience, Heidelberg, Germany).
4.9. Chromium Release Assay
The cytolytic capacity of CAR-RNA-transfected T cells was determined in a standard chromium release assay [50
]. Briefly, A375M and T2 cells were labeled with 100 µCi of Na251
cells. Target cells were washed and subsequently cultured in 96-well plates (Thermo Fisher, Waltham, MA, USA) at 1000 cells/well. The T cells were added at the indicated effector:target ratios in the absence or presence of different kinase inhibitors (as indicated), either alone or in combination. Cells were co-incubated in triplicate culture wells for 4–6 h. To determine spontaneous background release, target cells were incubated with R10 medium, whereas target cells cultured with 1% Nonidet-40 were used to determine maximum release. Radioactivity in the supernatant was determined and lysis was calculated as follows: ((measured release − background release)/(maximum release − background release) × 100%).
4.10. Figure Preparation and Statistical Analysis
Graphs were created and statistical analysis was performed using GraphPad Prism, Version 7 (GraphPad Software, La Jolla, CA, USA). p-Values were analyzed using a paired Students t-test. * indicates p ≤ 0.05, ** indicates p ≤ 0.01, and *** indicates p ≤ 0.001.